The goal of the proposed research is to gain further insight into processes involved in second language acquisition. The first aim is to investigate the effects of type of learning (rule-based vs. associative), age of learning, and their interaction. Our second aim is to investigate the role of first language transfer in the formation of rules in a second language. Our third aim is to investigate the ability to acquire production proficiency of stress accent by learners whose native language exhibits non-stress accent or tone. The results and interpretation of the proposed behavioral studies will provide a theoretical and empirical background for further investigation with neuro-physiological measures. Insight into the mechanisms of language acquisition and their interaction with age of acquisition can contribute to advancements in the development of second language pedagogy and in linguistic rehabilitation after strokes. In the proposed project, the acquisition of English stress, both rule-governed and memorized, will be investigated. Speakers of a variety of first languages will be included in the research. Native speakers of Japanese (a language that specifies accent lexically), French (a language that has predictable rule-governed accent), Spanish (a language that has both rule-based and lexically specified accent), and Thai (a language that has lexically specified tone) will participate. Native English controls will also be tested. Both early and late bilinguals will be included in order to investigate the application that, in the phonological domain, some types of learning (rule-based) are subject to a critical period whereas other types of learning (associative or memorized) are possible throughout the life span. The experimental work will be carried out over a two- year period. Fall and winter of 2001 will be spent on pilot work and protocol design. Subjects will be run the following spring, summer, fall and winter. During the spring and summer of 2003, the data will be analyzed and the results written up and submitted as journal articles. In one experiment, the acquisition of English stress placement rules will be investigated. Subjects will be asked to concatenate aurally presented syllables into nonsense words exhibiting noun, verb, and derivational stress patterns. In another experiment, the treatment of rule-governed and lexically accented (memorized) real English words will be investigated. Finally, the acquisition of stress accent production characteristics will be investigated. Acoustic measures of duration, fundamental frequency, and amplitude as well as physical measures of airflow and air pressure will be compared in stressed vs. non-stress syllables in minimal pairs such as permit(accent mark over the e)/permit. The data collected in these experiments will be analyzed for an effect of age of acquisition as well as an effect of accent type (rule-governed vs. lexicalized). Effects of native language phonology will also be investigated in cross-linguistic comparisons.